A few weeks ago, I wrote a summary of the website I have redesigned for my client, Brian McKenna, who created SnorEraser. SnorEraser is the only product that is 100% guaranteed to neutralize 100% of snoring noise, and the way it works is truly remarkable, far better than other snoring aids on the market.
I thought I’d give you a brief update, because the campaign I launched with did NOT deliver the immediate positive impact I anticipated.
In fact, where Brian’s previous site had converted an average of 2.66% of visitors into buyers since the start of the year (2015), my redesigned process was converting at only 0.87% (second half of August)!
The good news is that I have been able to make some quick changes, and the site is now converting at over 2 percent. More importantly, the average revenue per visitor is more than 6% up on the previous site.
It’s really important that I share this experience with you, because it can deliver some vital marketing lessons…
- No one’s “first best guess” is likely to be optimal.
- “Best practice” sometimes isn’t.
- Some factors can make a huge difference.
I’ll take you through these three lessons in some detail.
Lesson 1: No one’s “first best guess” is likely to be optimal
I’ve written at length about the “first best guess” approach before, particularly in “Convert!” arguing that real marketing should respond to the reality of what’s out there.
The fact is, we don’t know how the market is going to respond to any offer. There are so many factors that can affect the way people think, respond, and act that it’s impossible to predict with any accuracy.
In the case of SnorEraser, I made a number of assumptions in my campaign design…
The major assumption was that it would be better to show people how well SnorEraser works, by making a “free trial download” instead of pushing straight for the sale.
Lesson 2: “Best practice” sometimes isn’t.
Now, that assumption makes complete sense. It’s logical. It made for good reading in my campaign design.
Surely it’s easier to get someone to enter their email for a free trial than it is to get them to whip out their credit card, right?
And then, surely, once they try it and realize it actually works (which it does), they’ll be really keen to buy — particularly when we’re dealing with a problem that has real pain associated with it, and no truly effective alternative.
But I believe I was wrong.
In my seminars on Conversion Rate Optimization, I like to give people my list of essential CRO rules. It isn’t a long list.
That’s it. If there’s only one thing you need to know, it’s that you don’t know jack!
That actually frees you up not to have to get it right first time — and that changes everything!
Lesson 3: Some factors can make a huge difference.
One of the problems with the “first best guess” redesign approach is that you’re changing everything!
Very often, you’re changing most of the things about the old site that don’t work, and you’re changing most of the things that work as well.
When I complete redesigned SnorEraser.com, I changed the copy, the graphic design, the videos, everything. The question we have to ask is, which of those factors have the greatest impact on conversions?
Here’s the old home page (above the fold). Click to enlarge.
And here’s my (initial) redesign. Click to enlarge.
The key here is that, of all the improvements I think I made (better copy, better typography, simpler navigation, videos, etc. etc.), I also made a critical change — based on an assumption — that had far more of an impact than all those minor changes combined!
My assumption was that the “Free Trial” approach would get more sales than the “Buy Now” approach. How wrong that was.
In reality, just under 5% of visitors signed up for the free trial. That meant I would need more than 50% of those people to use the free trial and then return to the site to buy, even to match the previous conversion rate of 2.66%.
So, after the first few weeks of running the new site, when we were convinced that conversions were way down on the old site, we had to ask the following questions:
- What assumptions have we made?
- What big changes have we made, based on those assumptions, which might have impacted conversions?
- How can we quickly test to find out?
This is part of the core approach of the Lean Startup model. Lean Startup is a process for rapidly launching a business idea as cost-effectively as possible. It specifically sets out to test assumptions as early as possible, and as easily and cheaply as possible, to avoid the situation of investing a lot of money in ideas that may have a fundamental flaw built-in.
Here’s how I think my logic fell down…
- I assumed that Brian’s “origin story” video would be as engaging for new visitors to SnorEraser as it was for me when we first started talking about the project.
- I don’t know whether that is the case, but changing the videos on the home page was quite a big change, so we ended up re-using the more direct videos that were on the old home page, and moving Brian’s story onto the About page.
- I assumed that significantly more people would sign up for the free trial than buy immediately.
- I was correct that more people would sign up, but the rate was less than double the previous conversion rate. That is not significantly more.
- I assumed that the majority of people who signed up would download and try the free trial.
- In fact, we know that 76% of the people who signed up visited the free trial download page. That may seem good, but it also shows that we may have lost a quarter of the people who were interested enough to sign up!
- I assumed that, once people tried the free trial, they would be highly likely to come back and buy the product.
- The reality was that just 35% of free trial subscribers ended up becoming buyers, even with a follow-up sequence that would remind them a few times about the product over the following couple of weeks.
To summarize, I think the initial “Free Trial” assumption was wrong.
On the old site, 2.66% of visitors bought, despite all its apparent problems.
My redesigned funnel worked like this:
- 4.8% of visitors signed up
- Of those, 76% visited the trial download page
- Of those, 35% became buyers
That’s 4.8% × 76% × 35% = 1.27%
Less than half the original conversion rate!
An important factor I failed to take into account was simply that SnorEraser (currently) only costs $39!
$39 puts it into the impulse-buy category. Most people can throw $39 at a possible solution.
Maybe a $39 price tag just suits a “buy now” approach? Maybe people are happy to buy now at that price? Maybe they want the solution NOW, without having to go through a free trial? Maybe they don’t want to sit through a video at all?
(If the investment were $139, it might be different. It’s quite possible that, at a higher price, a longer sales process would be more appropriate.)
Of course, all those are assumptions. So what I has to do (following the lean approach) was to TEST.
I created another radically alternative home page.
- This page has two “Buy Now” buttons, and NO “Send My Free Trial” call to action.
- It also does not show the slide-up “Send My Free Trial” footer CTA.
- Instead of focusing the visitor on Brian’s (fascinating) story, we just pushed the core features you need for a sale: a powerful promise to solve your problem, backed up with all the reasons why you can trust us, including their unmatched 90-day guarantee.
The result is a much more direct-selling home page. It only took me a couple of hours to create, using Thrive Content Builder. And that’s exactly the right way to work, because my purpose here was to test an assumption — not to try to deliver perfection!
Since I put the new page live on September 7th, the conversion rate is back to over 2%, but the average revenue per visitor is higher than it was on the previous website.
Plus there are other valuable benefits. In the client’s words, because people are now clearer about what the product is…
Okay, here’s the low-down. No matter how experienced you are in marketing, you don’t know jack! You cannot predict what is going to work, and what isn’t.
(By the way, that’s why the people who claim to be able to offer you a foolproof 6-figure out-of-the-box no-lose cookie-cutter swipe campaign are full of shit! Like Ryan Deiss, for example, or Ryan Deiss.)
What does that tell us about marketing? It’s that marketing is not about getting it right first time. There is no best practice.
Well, maybe there is. It would have to be this…
Marketing is about having ideas, testing, learning, and responding, based on the facts.
And because online publishing gets cheaper, easier, and quicker every month, that’s great news for us!
That’s why I’m putting all my energy right now into developing a new revenue-share service model for web designers and marketing professionals.
There’s a bunch of compelling reasons why we could all benefit by moving to a model where we only get paid when our work proves it works. One of those is that it compels us to think carefully about the fastest route to discovering what works, and following that through to the conclusion.
You know, under the old model (the one where the client pays a significant fee, and has to accept the result), Brian McKenna would have been truly stung! He would have wasted his money. Fortunately, that’s not the way I work these days. I’m still working on SnorEraser, to make Brian significantly MORE money than he was making before. Because, when I do that, I’m also likely to get recurring residual income too. It’s a win-win, and it’s very exciting!
I could write all day about how perfect this model could be, but first I’m going to spend a few months figuring it out and proving it. So watch this space 🙂